Research Interest

Complete publication list can be found the here .

Survival Analysis and Semiparametric Modeling
  1. Forster AM, Tian L, and Wei LJ. Estimation for the Box-Cox Transformation Model Without Assuming Parametric Error Distribution. Journal of the American Statistical Association (JASA), 2001; 96: 1097-1101.
  2. Tian L, Zucker D, and Wei LJ. On the Proportional Hazards Model with Time-varying Regression Coefficients. Journal of the American Statistical Association (JASA), 2004, 100:172-183.
  3. Cai T, Tian L, and Wei LJ. Semiparametric Box-Cox Power Transformation Models for Censored Survival Observations, Biometrika, 2005, 92: 619-632.
  4. Park Y, Tian L, and Wei LJ. One- and two-sample Nonparametric Inference Procedures in the Presence of a Mixture of Independent and Dependent Censoring. Biostatistics, 2006; 7: 268-285.
  5. Tian L and Cai T. On Accelerated Failure Time Model for Current Status and Interval Censored Data, Biometrika, 2006; 93: 329-242.
  6. Tian L and Lagakos S. Analysis of a Partially Observed Binary Covariate Process and a Censored Failure Time in the Presence of Truncation and Competing Risks, Biometrics, 2006; 62: 821-828.
  7. Tian L and Huang J. Two-part Model for Analyzing Medical Cost Data in the Presence of Right Censoring, 2007, Statistics in Medicine, 23:4273-4292.
  8. Tian L and Wei LJ, Discussion on “Maximum likelihood estimation in semiparametric regression models with censored data”, JRSSB, 2007, 69:557.
  9. Tian L, Cai T, Zhao L and Wei LJ, On the Covariate-adjusted Estimation for an Overall Treatment Difference with Data from a Randomized Comparative Clinical Trial. Biostatistics, 2011, in press.
  10. Seok J, Tian L, Wong WH. Density Estimation on Multivariate Censored Data with Optional Pólya Tree. Biostatistics. 2014 Jan;15(1):182-95.
  11. Parast L, Tian L, Cai T. Landmark Estimation of Survival and Treatment Effect in a Randomized Clinical Trial. J Am Stat Assoc. 2014 Jan 1; 109(505):384-394.
  12. Parast L, Cai T, Tian L. Evaluating Surrogate Marker Information using Censored Data. Stat Med 2017; 36(11):1767-1782.
  13. Fei, J., Tian L. and Wei LJ. Robust Alternatives to ANCOVA for Estimating the Treatment Effect via a Randomized Comparative Study. Am Stat Assoc. 2018 (in press).
Restricted Mean Survival Time

The difference (or other contrasts such as ratio) in restricted mean survival time (RMST) offers an appealing alternative to hazard ratio: as the expected life expectancy within a time window, RMST is more interpretable for clinicians and patients; as the area under the survival curve, RMST is easy to present graphically; as a functional of the survival function, it can be estimated. I am interested in developing a whole set of statistical methods to promote its use in clinical trials. Currently, I am working on the associated covariates adjustment method, the objective selection of the truncation time point in RMST and generalization to recurrence event and competing risk settings.

  1. Tian L, Zhao L, Wei LJ. Predicting the restricted mean event time with the subject's baseline covariates in survival analysis. Biostatistics. 2014 Apr; 15(2):222-33.
  2. Uno H, Claggett B, Tian L, Inoue E, Gallo P, Miyata T, Schrag D, Takeuchi M, Uyama Y, Zhao L, Skali H, Solomon S, Jacobus S, Hughes M, Packer M, Wei LJ. Moving beyond the hazard ratio in quantifying the between-group difference in survival analysis. J Clin Oncol. 2014 Aug 1; 32(22):2380-5.
  3. Uno H, Wittes J, Fu H, Solomon SD, Claggett B, Tian L, Cai T, Pfeffer MA, Evans SR, Wei LJ., Alternatives to hazard ratios for comparing the efficacy or safety of therapies in noninferiority studies. Ann Intern Med. 2015 Jul 21; 163(2):127-34.
  4. Zhao L, Claggett B, Tian L, Uno H, Pfeffer MA, Solomon SD, Trippa L, Wei LJ., On the restricted mean survival time curve in survival analysis. Biometrics. 2016 Mar; 72(1):215-21.
  5. Cronin A., Tian L. and Uno H. strmst2 and strmst2pw: New commands to compare survival curves using the restricted mean survival time. The Stata Journal, 2016; 16(3):1-15.
  6. Yong F., Tian L., Yu S., Cai T., and Wei, L.J., Optimal stratification in outcome prediction using baseline information, Biometrika 2016; 103(4):817-828.
  7. Pak K, Uno H, Kim DH, Tian L, Kane RC, Takeuchi M, Fu H, Claggett B, and Wei LJ. Interpretability of cancer clinical trial results using restricted means survival time as an alternative to hazard ratio. JAMA Oncol. 2017; 3(12):1692-1696.
  8. Tian L, Fu H, Ruberg SJ, Uno H and Wei LJ. Efficiency of two sample tests via the restricted mean survival time for analyzing event time observation. Biometrics 2017; doi: 10.1111/biom.12770.
  9. Zhao L, Tian L, Claggett B, Pfeffer M, Kim DH, Solomon S and Wei LJ. Estimating treatment effect with clinically interpretation from a comparative clinical trial with an endpoint subject to competing risks, JAMA Cardiol. 2018; doi:10.1001/jamacardio.2018.0127.
  10. Huang B, Tian L, Talukder E, Rothenberg M, Kim DH, and Wei LJ. Evaluating treatment effect based on duration of response for a comparative oncology study. JAMA Oncology 2018 (in press).
  11. Horiguchi M., Tian L, Uno H., Cheng S, Kim DH, Schrag D and Wei LJ. Quantifying the long-term survival treatment effect in a comparative oncology clinical study. JAMA Oncology 2018 (in press).
  12. Uno H, Claggett B, Tian L, Fu H, Huang B, Kim DH, and Wei LJ. Adding a New Analytical Procedure with Clinical Interpretation in the Tool Box of Survival Analysis. Ann Oncol. 2018 Apr 3. doi: 10.1093/annoc/mdy109.
  13. Claggett, B., Tian L., Fu, H., Solomon, S., and Wei, LJ. Quantifying the totality of treatment effect with multiple event-time observations in the presence of a terminal event from a comparative clinical study. Biostatistics 2018 (in press).

Resampling Method

  1. Goldwasser MA, Tian L, and Wei LJ. Statistical Inference for Infinite Dimensional Parameters via Asymptotically Pivotal Estimating Functions. Biometrika, 2004; 91: 81-94.
  2. Tian L, Liu J, Zhao M, and Wei LJ. Statistical Inferences Based on Non-smooth Estimating Functions. Biometrika, 2004, 91:943-954.
  3. Tian L, Liu J, and Wei LJ. Implementation of Estimating Function Based Statistical Inferences via MCMC Samplers (with Discussion), JASA, 102:881-900, 2007
  4. Minnier J., Tian L and Cai T. A Perturbation Method for Inference on Regularized Regression Estimates. JASA, (2012). 106(496):1371-1382.

Meta Analysis

Meta analysis is probably the most used and also the most mis-used statistical method in clinical research. The statistical inference method associated with the random effects model used in meta analysis requires many assumptions, whose validity is often neglected in practice. For example, the number of studies needs to go to infinity in theoretical justifications, but the relevant method is used for analyzing data with even 3 or 4 studies. You can hardly call n=4 is big enough to allow asymptotic inference. A common type of meta analysis is combining ORs from multiple two by two tables. When some of the tables are sparse, then observed ORs don’t follow normal distribution. But the normal-normal random effects model is often still used due to the lack of alternatives. I am aiming to develop a set of statistical methods offering exact inference results without relying on those stringent assumptions.

  1. Tian L, Cai T, Pfeffer M, Piankov N, Cremieux P and Wei LJ. Exact and Efficient Inference Procedure for Meta-analysis And Its Application to the Analysis of Independent 2 x 2 Tables With All Available Data But Without Artificial Continuity Correction. Biostatistics, 10:275-281, 2009
  2. Wang R, Tian L, Cai T and Wei LJ, Nonparametric Inference Procedure for Percentiles of the Random Effect Distribution in Meta Analysis. Annals of Applied Statistics, 4:520-532, 2010
  3. Claggett B, Minge Xie and Tian L. Meta Analysis with Fixed, Unknown, Study-specific Parameters. Biostatistics. J Am Stat Assoc. 2014, 109:1660-1671.
  4. Hasegawa T., Claggett B., Tian L., and Wei L.J. The Myth of Making Inferences for an Overall Treatment Effect with Data from Multiple Comparative Studies via Meta-analysis and Beyond. Statistics in Biosciences 2016; 9(1):1-14.
  5. Liu S, Tian L., Lee S and Xie MG. Exact Inference on Meta Analysis with Generalized Fixed Effects and Random-effects Models. Biostatistics & Epidemiology 2018; 1(2):1-22.
  6. Wang Y and Tian L. An Efficient Numerical Algorithm for Exact Inference in Meta analysis. Journal of Statistical Computation and Simulation 2018; 88:646-656.
  7. Michael, H., Thornton, S., Xie, M., and Tian L. (2018) Exact Inference on the Random Effects Model for Meta Analysis with Few Studies. Biometrics (submitted).
  8. Gronsbell, J., Nie, L., Ying, L. and Tian L. (2018) Exact Inference for the Random Effects Model for Meta Analyses with Rare Events. Statistics in Medicine (submitted).

High Dimensional Data Analysis

  1. Park PJ and Tian L. Linking Gene Expression Data with Patient Survival Time Using Partial Least Squares. Bioinformatics, 2002; 18: 120S-127S.
  2. Xu X, Tian L, and Wei LJ. Combining Tests for Linkage or Association across Multiple Phenotypic Traits. Biostatistics, 2003:4: 223-229.
  3. Zhong S, Tian L, Li C, Storch FK, and Wong WH. Comparative Analysis of Gene Sets in the Gene Ontology Space under the Multiple Hypothesis Testing Framework. Proc. IEEE Comp Systems Bioinformatics. 2004, 425-435.
  4. Tian L, Greenberg S, Kong SW, Altschuler J, Kohane IS, and Park P. Discovering Statistically Significant Pathways in Expression Profiling Studies, Proceedings of National Academy of Sciences (PNAS), 2005, 102: 13544-13549.
  5. Cai T, Huang J, and Tian L. Regularized Estimation for the Accelerated Failure Time Model, Biometrics, 65:394-404, 2009.
  6. Zhao XG, Yi L, Dai W, and Tian L, AUC based Biomarker Ensemble with an Application on Gene Scores Predicting Low Bone Mineral Density, Bioinformatics, 2011, 27(21): 3050-3055.
  7. Tian L, A Alizadeh, A Gentles and Tibshirani R, A simple method to estimating the interaction between a treatment and a large number of covariates, J Am Stat Assoc. 2014, 109(508):1517-1532.
  8. Zhang Z, Ying L and Tian L. On feature ensemble optimizing the sensitivity and partial ROC curve. Statistca Sinica 2018 (in press).
  9. Tian L., Liu, Y., Fire, A., Boyd, S. and Olshen, R. Clonality: Point estimation. The Annals of Applied Statistics 2018 (in press).

Statistical Methods for Personalized Medicine

Precision medicine becomes a hot “buzz” word today. From the statistical perspective, its simplest version still is about the treatment covariate interaction detection. While AI, outcome learning, dynamic optimization, machine learning, et al can all be useful tools for this purpose, the basic idea is still trying to best detect the interaction, while bypassing the need of predicting individual’s outcome under treatment of interest or standard care. I am interested in (1) how to identify the best subgroup of patients, who may have the most promising response to the treatment; (2) how to validate and make statistical inference on the treatment effect in the selected subgroup; (3) how to design an adaptive enrichment trial allowing the simultaneous detection of the promising subgroup and valid inferences for the corresponding treatment effect. Currently, I am moving towards the direction of harmonizing the information from RCT and observation studies for the development of precision medicine. More papers on the way!

  1. Tian L, Wang W, and Wei LJ. Estimating Predictors for Long- or Short-term Survivors. Biometrics, 2003; 59: 1008-1015.
  2. Uno H, Tian L, and Wei LJ. The Optimal Confidence Region for a Random Parameter, Biometrika, 2005, 92: 957-964.
  3. Uno H, Cai T, Tian L, and Wei LJ. Evaluating Prediction Rules for t-year Survivors with Censored Regression Models, JASA, 2007, 102:527-537.
  4. Tian L, Cai T, Goetghebeur E, and Wei LJ. Model Evaluation Based on the Sampling Distribution of Estimated Absolute Prediction Error, Biometrika, 2007; 94: 297-311.
  5. Cai T, Tian L, Solomon S, and Wei LJ. Predicting Future Responses based on Possibly Misspecified Working Models, Biometrika, 95: 75-92, 2008.
  6. Tian L, Cai T, and Wei LJ. Identifying Subjects Who Benefit from Additional Information for Better Prediction of the Outcome Variables, Biometrics, 2008: 10.1111/j.1541-0420.2008.01125.x
  7. Cai T, Tian L, Uno H, Solomon S, and Wei LJ, Calibrating Parametric Subject-Specific Risk Estimation. Biometrika, 2010 97(2):389-404.
  8. Tian L and Tibshirani R, Adaptive Index Model for Marker-based Risk Stratification, Biostatistics 2011, 12(1): 68-86.
  9. Tian L, Cai T, R. Wang and L.J. Wei, The Highest Confidence Density Region and its Usage for Joint Inferences About Constrained Parameters, Biometrics, 2011 67(2):604-610.
  10. Li Y, Tian L and Wei LJ, Estimating Subject-Specific Dependent Competing Risk Profile with Censored Event Time Observations, Biometrics, 2011 67(2):427-35.
  11. Cai T and Tian L, Analysis of Randomized Comparative Clinical Trial Data for Personalized Treatment Selections. Biostatistics, 12(2):270-282.
  12. Uno H, Cai T, Tian L and Wei LJ, Graphical Procedures for Evaluating Overall and Subject-specfic Incremental Values from New Predictors with Censored Event Time Data. Biometrics, 2011, in press.
  13. Zhao L, Tian L, Cai T, Claggett B, Wei LJ. Effectively selecting a target population for future comparative study. J Am Stat Assoc. 2013 Jan 1; 108(502):527-539.
  14. Delmar, P., Cornelia, I., and Tian L. Innovative methods for the identification of predictive biomarker signature in oncology: application to Avastin. International Journal of Clinical Trials 2017; 5:107-115.
  15. Huang X, Sun Y, Chatterjee S, Chakravartty A, Tian L, and Devanarayan V. Patient subgroup identification for clinical drug development. Statistics in Medicine 2017; 36(9):1414-1428.
  16. Michael H, Tian L. and Ghebremichael M. The ROC curve for regularly measured longitudinal biomarkers. Biostatistics 2018; doi: 10.1093/biostatistics/kxy010.
  17. Chen S, Tian L, Cai T, and Yu M. A general statistical framework for subgroup identification and comparative treatment scoring. Biometrics 2017; 73(4):1199-1209.